PROJECT SUMMARY/ABSTRACT Homeless adults are 8 times as likely to be alcohol dependent compared with adults in the general population, yet few studies have examined the precipitants of alcohol use in this vulnerable population. This lack of information is a barrier to developing effective alcohol treatments for homeless adults, a population that is often excluded from clinical trials. Ecological momentary assessments (EMAs) that involve repeated assessment of thoughts/mood/behaviors (e.g., via smart phone) is currently the most accurate way to assess individuals in real-time in their natural environments. Advances in smartphone technology also allow for the collection of continuous geolocation and other passive sensing data. Thus, researchers can now link environmental risks and protective factors to outcomes, without reliance on subjective reporting alone. Building on our prior work, we propose a three-phase study to develop and test a ?just in time? adaptive intervention to reduce alcohol use in homeless men and women. This application is submitted in response to PA-16-073, which encourages the development of innovative behavioral treatments for alcohol use. During Phase I, we will use smartphones and passive sensing technologies to monitor geolocation, psychosocial variables (e.g., stress, affect, urge to drink), and alcohol use in a group of 80 homeless adults enrolled in shelter-based substance abuse treatment. We will identify environmental (i.e., geolocation), cognitive, and behavioral antecedents of alcohol use over 4 weeks. In Phase II, we will use this information to create a risk algorithm and tailored treatment messages that anticipate and intervene to prevent drinking. We will modify an existing platform, previously validated for smoking cessation, to create the alcohol intervention. The app will assess imminent risk of alcohol use after each EMA and will deliver relevant treatment messages that match a person's current risk factors. Other app features will include ?Tips? and ?Resources? functions that will enable individuals to access treatment messages and other resources on demand (e.g., ways to enhance motivation, coping with stress, refusal skills). In Phase III, we will test the feasibility, acceptability and preliminary efficacy of the app in a sample of 40 homeless adults enrolled in shelter-based treatment who receive the EMA plus treatment messages over 4 weeks. Drinking will be determined via self-report, supplemented by a transdermal alcohol sensor (i.e., SCRAM) worn by a random half of participants. This project will be the first to combine geolocation and psychosocial variables to identify real-time antecedents of drinking. If effective, this smartphone app could significantly improve treatment engagement, drinking outcomes, and quality of life among homeless adults with alcohol use disorders.